Identification of rabbit immunoglobulin latent Cκ1 allotype genes alters the concept of allelic inheritance

Abstract

In the rabbit, expression of immunoglobulin Cκ1 light chain genes is believed to be under allelic control. Conventionally, four nominal allotypic variants, b4, b5, b6 and b9 have been shown to be co-dominantly expressed at the Cκ1 gene locus. Analogously, the heavy chain allotypes, VHal, VHa2 and VHa3, found in the V region, are also believed to be inherited co-dominantly. However, after our earlier discovery of non-allelic or latent allotypes in the serum and on cell surfaces, we subsequently reported that cDNA sequences for latent b5 and b6 were identical to nominal b5 and b6, respectively (Ishaq et al., 1990). The latent b5 cDNAs were from two homozygous b4,b4 rabbits; the latent b6 cDNA was found in a heterozygous b4,b9 rabbit. The cDNA sequences had been obtained from lymph nodes and spleens of rabbits which had been infected with Trypanosoma brucei in order to induce latent allotypes more consistently. In this article, employing spleen DNA from three different T. brucei-infected rabbits, (one, heterozygous b4,b9; two others, homozygous, b4,b4), we initially detected two bands by Southern analysis after Hind III digestion using a 624 base pair Cκ1 b4 probe derived from a b4,b4 rabbit. However, the probe was non-specific allotypically as it hybridized to b5, b6 and b9 Cκ1 DNA. Therefore, in order to search for the latent genes, we used allotype-specific oligonucleotides for b5, b6 and b9 to probe DNAs from both normal and T. brucei-infected rabbits by Southern blotting. At the outset, employing a b4 oligonucleotide probe, we detected a single 5.8 Kb segment in two b4,b4 rabbit DNAs after Bgl II digestion. The findings, using the 624 base pair Cκ1 b4 probe and the b4 oligomer, agreed with earlier data reported by others. Subsequently, we tested kidney, liver and spleen DNAs from one of these and other rabbits for genomic latent b5, b6 and b9 using these specific oligomeric probes. For each latent allotype, Southern analysis revealed latent-allotype specific DNA segments in the genome. After cosmid cloning and sequencing, latent κ1, b5, b6 and b9 genes were found to be identical in their coding regions with their nominal counterparts. The genes contained at the 5′ end the PyPyXPyAG RNA splice acceptor site found in immunoglobulin and many eukaryotic genes, as well as the termination codon TAG, together with AATAAA and the T-rich sites responsible for cleavage-polyadenylation in the untranslated region downstream from the 3′ end. Single cosmid clones representing the b5, b6 and b9 genes were mapped for restriction sites which resulted in identifying putative Jκ and enhancer regions. The results thus indicate that latent allotype genes are potentially functional. The data provide evidence that allotypes are not strictly controlled by allelic genes but must be regulated by an hierarchical mechanism which provides for synthesis of allelic allotypes mainly (10–20 mg/ml) together with non-allelic allotypes at lower concentrations (2–20 μg/ml) following activation of the latent genes. These results lay to rest the belief that Cκ1 latent allotypes are the products of scrambled genes or idiotypic mimicry. Importantly, we now have the possibility of investigating the factors leading to latent allotype gene expression, the Vκ and Jκ regions associated with the genes, and therefore whether antibody diversity is expanded. We do not know, nor do we imply, that latent allotypes are present in all rabbits. However, since the four conventional Cκ1 allotypes are present in the genome of several of our tested rabbits, and are presumably functional, we are faced with the probability that rabbit allotypes under certain conditions may in fact behave as isotypes and not allotypes.